Can You Figure Out What's Wrong with These Alaska Crime Report Charts?

Alaska's CRIME IN ALASKA 2017:  Uniform Crime Reporting Program Annual Report came out today.  You can get to the whole report  here.  I haven't read through the whole thing, because the first chart was so bad.  I got distracted looking at what was wrong with this chart.

Charts are supposed to help visual words or numbers, help you see relationships the column of numbers or verbal descriptions the visualize.  So here's the first chart in the report.

Look at this chart.  Can you tell what the numbers represent?  The gold (Non-Reporting) if I understand this right, is the number of people in jurisdictions that do not report crime statistics.  That makes sense because the population in 2017 was 739,000 people.

That is a problem, but not the major problem I see here.

Can you tell what my real problem is?  I'm going to distract you with some pictures that have nothing to do with this so you can think and not be distracted by my corrected chart.  Think about jeopardy music playing while you look at the pictures.

The raspberries are ripening daily in the back yard.













Have you figure out what's wrong with the chart?  If yes, scroll down.  If not, go back and figure it out.

A while back I posted about the city turning off the water one day because they're doing work on the old pipes.

Well, there are some upsides to this.  They're putting in new sidewalk/bike path along Lake Otis. Unfortunately, I think it's not related to which parts need it the most, but rather to where they had to dig to put in new pipes.  But it's better than nothing.  It was pretty gnarly.



OK, now go look at the chart I made of these numbers.  (I'm afraid I got green instead of gold for the non-reported population. I'm doing this in Mac's Numbers and it has limited combinations of colors.)

Here's a more accurate representation of the numbers:

As I said, the point of a chart is to help visualize the meaning of the data.  In the State report's chart, the vertical columns segment out the two different populations, but the length of the blue and gold has no relationship to the numbers.  The gold for 2013 is about 40% of the chart, when the actual number of unreported population is less than 1%.  The same distortion exists for all five years.   And the difference in total population for each year is also exaggerated.  In fact, the differences are so small that in a proportionally correct chart, you can barely see the gold (green in my chart) or the differences between the years.  They're really insignificant.  

Here's another unrealized chart.  (I'm trying to be diplomatic here.)  

The chart is better than trying to find this information in the text, but it could have organized the information in much more useful ways.  

My biggest problem is that I can see no order in the crimes selected.  The point seems to be to graphically show us how frequently different crimes occur.  So why not put them in order from least frequent to most frequent?  

The spacing around a clock isn't really relevant to anything.  The clock tells us nothing except this is about time.  But why is ROBBERY at the 3?  All the critical information is presented in words, not images.  As I said, all this does (and this is not unimportant) is highlight these frequencies by pulling them out of the text of the report and putting them in an image.  But the image itself doesn't show much thought about how to convey the numbers visually.  

Data visualization is one of the hottest fields in journalism today.  DATAVIZ folks are taking vast amounts of data and putting them into great charts that help people see complex ideas.  Here are some ideas on this from one of the leaders in the field, Alberto Cairo.

But these examples take Stone Age data charts and botch them.